在本論文中,我提出對表述問題的一條新進路。我將表述問題重新界定為以下的兩個問題:(一)是何依據使得我們能單義地使用同一個述詞來表述不同事物?(二)是什麼能整合一個句子的各部份(通常為一名稱與一述詞),成為一個單一的表述,而非僅僅為字詞的表列?我們可以稱第一個問題所關心的為「分類的單一性」,第二個為「命題的單一性」;一個表述理論,便是對這兩個問題所提出的系統性回答。 前三章我將批評三種對表述問題的傳統進路:實在論、惟名論、以及概念論。實在論者解釋表述的方式,是去設置某些抽象的元目做為述詞的語義值,並試圖證明表述只不過是在個體與抽象元目之間的一種關係。我將特別討論殊質理論,尤其是培根(1995)新近的陳述與討論。在第二章,我把惟名論的進路描繪如下:他們試圖透過將我們所使用的語言,翻譯或重述為惟名論可接受的語言,作為表述的語言基礎。我將討論卡納普(1928)從感覺與料建構性質集的計劃,與費爾德(1980)惟名論化科學理論的計劃,將之應用在處理表述問題上的可能性。至於概念論的進路,我將追溯此傳統至康德對概念的超驗分析,與布倫坦諾的意向性,但我的重點會放在麥農的對象理論,和假裝理論,兩個理論上。我將論證,以上的三種進路都注定要失敗。 我把傳統進路失敗的原因分析為分裂的問題:它們皆試圖透過劃分來為表述放下基礎,並努力將表述的本質化約至此基礎上。因此它們無法認知到,在一個表述理論當中,有三個因素是不可或缺的:世界與世界中事物之排列、語言的邏輯結構、以及心靈做判斷的能力。根據這樣的分析,我提出了一條新的進路──整體論式的進路,即,一個將這三個重要因素結合在一起的統一理論。我注意到蒯因和戴維森的整體論計劃,恰好足夠成為這樣一條進路的背景,因此我在第四章中將發展並應用他們的計劃來解釋表述。我的結論是:整體論式的進路,是惟一能夠對表述問題給出另人滿意的解釋之進路。
In this thesis I present what I believe to be a new approach to the problem of predication. I shall reformulate the problem of predication as the following two questions: (1) What justifies our univocal use of one and the same predicate expression to predicate of several different things? (2) What unifies the parts of a sentence, usually a name with a predicate, to make a single predication rather than a mere list of words? Let us call what the first question concerns “the unity of classification,” and the second “the unity of the proposition”; a theory of predication, then, is a systematic account to answer such two questions. In the first three chapters, I provide a critical review of three traditional approaches to predication in turn: the realist, the nominalist, and the conceptualist approach. The realist approaches predication by positing abstract entities as the semantic values of predicates, attempting to show that predication is nothing but a relationship between an individual and an abstract entity. Special attention will be paid to the Trope theory, especially J. Bacon’s (1995) new formulation and discussion. In Chapter Two, I characterize the nominalist approach as one that explains predication by some way of translation or reformulation of the language in use into a “nominalistically acceptable language” as the linguistic foundation for predication. I discuss the possibility of applying Carnap’s (1928) project of constructing quality classes from sense-data, and H. Field’s (1980) nominalist program of nominalizing the scientific theory, to an account of predication. As for the conceptualist approach, I trace this tradition back to Kant’s transcendental analysis of concepts and Brentano’s intentionality thesis, but my focus will be on the Meinongian theory of object and the pretense theory. I shall argue that the three aforementioned approaches are all doomed to fail. I analyze the failure of the traditional approaches as the cleavage problem: all of them attempt to lay some foundation for predication by making cleavages, and endeavor to reduce the nature of predication to such a foundation. They therefore fail to recognize that in a theory of predication, three factors are indispensable: the world with the arrangement of things in it, the logical structure of our language, and the mental faculty of the making judgments. Based on such an analysis, I propose a new approach to predication—the holistic approach, one which intends to combine the three important factors to yield a unified theory of predication. I notice that Quine and Davidson’s holistic programs happen to provide such a background sufficient for a full-scale theory of predication, which I shall work out in Chapter Four. I conclude that the holistic approach I advocate is the only possible approach that satisfactorily explains predication.